vllm
vllm copied to clipboard
Published
1 week ago
•
vllm-project
vllm-project
[Bug]: Potential memory leak: VRAM continuously increases and not freed with deepseek-r1 on vLLM v1 engine
Your current environment
The output of `python collect_env.py`
PyTorch version: 2.6.0+cu124
Is debug build: False
CUDA used to build PyTorch: 12.4
ROCM used to build PyTorch: N/A
OS: Ubuntu 22.04.4 LTS (x86_64)
GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
Clang version: Could not collect
CMake version: version 4.0.0
Libc version: glibc-2.35
Python version: 3.12.10 (main, Apr 9 2025, 08:55:05) [GCC 11.4.0] (64-bit runtime)
Python platform: Linux-5.15.0-136-generic-x86_64-with-glibc2.35
Is CUDA available: True
CUDA runtime version: 12.4.131
CUDA_MODULE_LOADING set to: LAZY
GPU models and configuration:
GPU 0: NVIDIA H20-3e
GPU 1: NVIDIA H20-3e
GPU 2: NVIDIA H20-3e
GPU 3: NVIDIA H20-3e
GPU 4: NVIDIA H20-3e
GPU 5: NVIDIA H20-3e
GPU 6: NVIDIA H20-3e
GPU 7: NVIDIA H20-3e
Nvidia driver version: 550.144.03
cuDNN version: Could not collect
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True
CPU:
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Address sizes: 52 bits physical, 57 bits virtual
Byte Order: Little Endian
CPU(s): 192
On-line CPU(s) list: 0-191
Vendor ID: GenuineIntel
Model name: Intel(R) Xeon(R) Platinum 8468V
CPU family: 6
Model: 143
Thread(s) per core: 2
Core(s) per socket: 48
Socket(s): 2
Stepping: 8
CPU max MHz: 3800.0000
CPU min MHz: 800.0000
BogoMIPS: 4800.00
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l3 cat_l2 cdp_l3 invpcid_single intel_ppin cdp_l2 ssbd mba ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb intel_pt avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local split_lock_detect avx_vnni avx512_bf16 wbnoinvd dtherm ida arat pln pts hwp hwp_act_window hwp_epp hwp_pkg_req avx512vbmi umip pku ospke waitpkg avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq la57 rdpid bus_lock_detect cldemote movdiri movdir64b enqcmd fsrm md_clear serialize tsxldtrk pconfig arch_lbr amx_bf16 avx512_fp16 amx_tile amx_int8 flush_l1d arch_capabilities
Virtualization: VT-x
L1d cache: 4.5 MiB (96 instances)
L1i cache: 3 MiB (96 instances)
L2 cache: 192 MiB (96 instances)
L3 cache: 195 MiB (2 instances)
NUMA node(s): 2
NUMA node0 CPU(s): 0-47,96-143
NUMA node1 CPU(s): 48-95,144-191
Vulnerability Gather data sampling: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Mitigation; Enhanced / Automatic IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI BHI_DIS_S
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected
Versions of relevant libraries:
[pip3] flashinfer-python==0.2.1.post2+cu124torch2.6
[pip3] numpy==2.2.4
[pip3] nvidia-cublas-cu12==12.4.5.8
[pip3] nvidia-cuda-cupti-cu12==12.4.127
[pip3] nvidia-cuda-nvrtc-cu12==12.4.127
[pip3] nvidia-cuda-runtime-cu12==12.4.127
[pip3] nvidia-cudnn-cu12==9.1.0.70
[pip3] nvidia-cufft-cu12==11.2.1.3
[pip3] nvidia-curand-cu12==10.3.5.147
[pip3] nvidia-cusolver-cu12==11.6.1.9
[pip3] nvidia-cusparse-cu12==12.3.1.170
[pip3] nvidia-cusparselt-cu12==0.6.2
[pip3] nvidia-nccl-cu12==2.21.5
[pip3] nvidia-nvjitlink-cu12==12.4.127
[pip3] nvidia-nvtx-cu12==12.4.127
[pip3] pyzmq==26.4.0
[pip3] torch==2.6.0
[pip3] torchaudio==2.6.0
[pip3] torchvision==0.21.0
[pip3] transformers==4.51.3
[pip3] triton==3.2.0
[conda] Could not collect
ROCM Version: Could not collect
Neuron SDK Version: N/A
vLLM Version: 0.8.4
vLLM Build Flags:
CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled
GPU Topology:
GPU0 GPU1 GPU2 GPU3 GPU4 GPU5 GPU6 GPU7 NIC0 NIC1 NIC2 NIC3 NIC4 NIC5 CPU Affinity NUMA Affinity GPU NUMA ID
GPU0 X NV18 NV18 NV18 NV18 NV18 NV18 NV18 PIX NODE SYS SYS SYS NODE 0-47,96-143 0 N/A
GPU1 NV18 X NV18 NV18 NV18 NV18 NV18 NV18 NODE NODE SYS SYS SYS NODE 0-47,96-143 0 N/A
GPU2 NV18 NV18 X NV18 NV18 NV18 NV18 NV18 NODE PIX SYS SYS SYS NODE 0-47,96-143 0 N/A
GPU3 NV18 NV18 NV18 X NV18 NV18 NV18 NV18 NODE NODE SYS SYS SYS NODE 0-47,96-143 0 N/A
GPU4 NV18 NV18 NV18 NV18 X NV18 NV18 NV18 SYS SYS NODE NODE PIX SYS 48-95,144-191 1 N/A
GPU5 NV18 NV18 NV18 NV18 NV18 X NV18 NV18 SYS SYS NODE NODE NODE SYS 48-95,144-191 1 N/A
GPU6 NV18 NV18 NV18 NV18 NV18 NV18 X NV18 SYS SYS NODE NODE NODE SYS 48-95,144-191 1 N/A
GPU7 NV18 NV18 NV18 NV18 NV18 NV18 NV18 X SYS SYS NODE NODE NODE SYS 48-95,144-191 1 N/A
NIC0 PIX NODE NODE NODE SYS SYS SYS SYS X NODE SYS SYS SYS NODE
NIC1 NODE NODE PIX NODE SYS SYS SYS SYS NODE X SYS SYS SYS NODE
NIC2 SYS SYS SYS SYS NODE NODE NODE NODE SYS SYS X PIX NODE SYS
NIC3 SYS SYS SYS SYS NODE NODE NODE NODE SYS SYS PIX X NODE SYS
NIC4 SYS SYS SYS SYS PIX NODE NODE NODE SYS SYS NODE NODE X SYS
NIC5 NODE NODE NODE NODE SYS SYS SYS SYS NODE NODE SYS SYS SYS X
Legend:
X = Self
SYS = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI)
NODE = Connection traversing PCIe as well as the interconnect between PCIe Host Bridges within a NUMA node
PHB = Connection traversing PCIe as well as a PCIe Host Bridge (typically the CPU)
PXB = Connection traversing multiple PCIe bridges (without traversing the PCIe Host Bridge)
PIX = Connection traversing at most a single PCIe bridge
NV# = Connection traversing a bonded set of # NVLinks
NIC Legend:
NIC0: mlx5_2
NIC1: mlx5_3
NIC2: mlx5_4
NIC3: mlx5_5
NIC4: mlx5_6
NIC5: mlx5_bond_0
NVIDIA_VISIBLE_DEVICES=GPU-0d0b33df-7e45-bfc6-f4c9-d0c4d902a0dd,GPU-18d384b3-b29e-d8e8-e14d-52bafd692141,GPU-1d57c269-e15c-f386-1e75-babffd220899,GPU-f182165b-4bc0-3cbf-9d6e-0ce03742941c,GPU-7e702a55-e829-8ae9-e041-4e7755d360a9,GPU-1a155106-f15d-4766-4058-6dc26678a3c0,GPU-0d072414-df92-4e72-25c1-28eb9296e1be,GPU-37e611b1-2128-d24a-36a1-77354d3cb2c5
NVIDIA_REQUIRE_CUDA=cuda>=12.4 brand=tesla,driver>=470,driver<471 brand=unknown,driver>=470,driver<471 brand=nvidia,driver>=470,driver<471 brand=nvidiartx,driver>=470,driver<471 brand=geforce,driver>=470,driver<471 brand=geforcertx,driver>=470,driver<471 brand=quadro,driver>=470,driver<471 brand=quadrortx,driver>=470,driver<471 brand=titan,driver>=470,driver<471 brand=titanrtx,driver>=470,driver<471 brand=tesla,driver>=525,driver<526 brand=unknown,driver>=525,driver<526 brand=nvidia,driver>=525,driver<526 brand=nvidiartx,driver>=525,driver<526 brand=geforce,driver>=525,driver<526 brand=geforcertx,driver>=525,driver<526 brand=quadro,driver>=525,driver<526 brand=quadrortx,driver>=525,driver<526 brand=titan,driver>=525,driver<526 brand=titanrtx,driver>=525,driver<526 brand=tesla,driver>=535,driver<536 brand=unknown,driver>=535,driver<536 brand=nvidia,driver>=535,driver<536 brand=nvidiartx,driver>=535,driver<536 brand=geforce,driver>=535,driver<536 brand=geforcertx,driver>=535,driver<536 brand=quadro,driver>=535,driver<536 brand=quadrortx,driver>=535,driver<536 brand=titan,driver>=535,driver<536 brand=titanrtx,driver>=535,driver<536
NCCL_VERSION=2.20.5-1
NVIDIA_DRIVER_CAPABILITIES=compute,utility
NVIDIA_PRODUCT_NAME=CUDA
VLLM_USAGE_SOURCE=production-docker-image
CUDA_VERSION=12.4.0
LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64
NCCL_CUMEM_ENABLE=0
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
CUDA_MODULE_LOADING=LAZY
🐛 Describe the bug
Bug Description
vLLM Version: v0.8.4 GPU: 8 x NVIDIA H20-3e Model: deepseekai/DeepSeek-R1 gpu_memory_utilization: 0.9 (default)
When sending chat requests with long input texts (>= 2048 tokens), initially using 32 concurrent requests, GPU memory increases but stays within a reasonable range (actual VRAM < Total VRAM * gpu_memory_utilization). However, after some time, when using 64 concurrent requests, the GPU memory rapidly grows to an alarming level (approximately 98% of Total VRAM, exceeding Total VRAM * gpu_memory_utilization). Furthermore, this extra memory is not released even some time after the requests have finished.
Steps to Reproduce
-
Start the vLLM server using the following command:
source /etc/profile && vllm serve /data/DeepSeek-R1 \ --port 8000 \ --served-model-name deepseekai/DeepSeek-R1 \ --api-key=YOUR_API_KEY \ --tensor-parallel-size 8 \ --max-model-len 163840 \ --max-num-batched-tokens 163840 \ --trust-remote-code \ --enable-prefix-caching \ --enable-reasoning \ --reasoning-parser deepseek_r1 -
Prepare a ShareGPT dataset with prompts longer than or equal to 2048 tokens using the provided script:
python3 filter_sharegpt.py \ --dataset-path /path/to/your/sharegpt.json \ --tokenizer deepseekai/DeepSeek-R1 \ --min-input-len 2048 \ --output /path/to/your/filtered_sharegpt.json -
Send requests using the provided script:
export OPENAI_API_KEY=YOUR_API_KEY # First run: Max concurrency 32 python3 send_req.py \ --model deepseekai/DeepSeek-R1 \ --tokenizer deepseekai/DeepSeek-R1 \ --base-url http://localhost:8000 \ --sharegpt-path /path/to/your/filtered_sharegpt.json \ --max-concurrent 32 \ --num-requests 128 # Wait for some time... # Second run: Max concurrency 64 python3 send_req.py \ --model deepseekai/DeepSeek-R1 \ --tokenizer deepseekai/DeepSeek-R1 \ --base-url http://localhost:8000 \ --sharegpt-path /path/to/your/filtered_sharegpt.json \ --max-concurrent 64 \ --num-requests 128
Expected Behavior
The actual GPU memory usage should not exceed Total VRAM * gpu_memory_utilization (i.e., 90% of total VRAM in this case). Memory allocated for requests should be released after processing.
Actual Behavior
GPU memory usage grows significantly across the two test runs (especially during the second run with higher concurrency), eventually exceeding Total VRAM * gpu_memory_utilization. The excess memory is not released after the requests are completed.
Before submitting a new issue...
- [x] Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.
